My daughter Noa and I recently wrote a song called “Bloomer.” It will not be among the most-streamed songs of the year. But it will make the top ten in our house.
The song isn't very good, but it's strangely popular because it's ours. We made it without knowing how to compose, play, or produce it. All we needed was an idea, a prompt, and Suno.
Most concern surrounding products like Suno has focused on the intellectual property used to train their models. That debate matters. But Suno also turns creation into a form of consumption: making a song becomes a way of choosing what to hear.
This shift toward just-in-time media—made on demand for a particular person and moment—will reshape media far beyond music.
From 1% to participation
Twenty years ago, Jakob Nielsen described participation inequality in online communities through the 90–9–1 rule. Roughly 90% of people observed, 9% contributed occasionally, and 1% accounted for most of the creation.
It was a rule of thumb, not a law. But it captured the early internet: almost everyone could publish, yet relatively few did. Watching was easier than making. Creating required opening a tool, starting with an idea, and deciding the result was worth sharing.
That imbalance survived social media. Instagram and YouTube gave almost anyone the ability to become a creator. But creating still required time, skill, and the motivation to make something for other people.
TikTok lowered the barrier further. Editing with CapCut is easier than with Final Cut. Features like Duets let people create by responding to what they just watched. But much of that participation starts with something already circulating: a trend, challenge, or piece of culture. Just-in-time media can start with a private joke, family ritual, or fleeting need. The result only has to matter to the person making it.
Nielsen’s recommendations now read like a description of generative media: make contributing easier, let people edit instead of facing a blank page, and make participation a side effect of something they already do.
That last point is the important one.
When Noa and I ask for a song based on her observations, are we making music or choosing what to hear? The request is both a creative instruction and a consumption decision. We do not have to become musicians. Creation is simply how we get the song we want.
The 1% rule will not disappear. Taste, skill, and ambition remain unevenly distributed. But the curve can bend, especially when what people make never needs an audience.
Media made for a moment
Most discussion of AI-generated music gravitates toward replacement and slop. Will a generated song take a stream from Taylor Swift? Will an endless supply of passable music make any individual track worth less?
Those questions matter. But just-in-time media will take occasions before it takes audiences. A brushing-teeth song. A joke between friends. A birthday soundtrack. Music timed to your workout. A bedtime story about the stuffed animal that went missing.
None needs to become a hit or even be shared to be meaningful. Its value comes from fitting one person, family, or moment.
Serving those moments will change the tools and platforms we use. The interesting question about Suno is not simply whether it can make good music. It is whether Suno is a creation tool, a consumption product, or an early sign that the distinction is becoming less useful.
Creation tools are opened with intent. Consumption products become habits. When creation moves inside those habits, people can make things without deciding to become creators. Making something becomes another way to get what they want to hear or watch.
The progression is already visible across media. Since ChatGPT arrived, we have used it to make personalized bedtime stories. Suno brings that behavior to audio. Video will follow, into what I’ve called promptable TV.
Today, music and television largely organize around a familiar division. Artists and studios create. Spotify and Netflix distribute. Audiences choose from a catalog.
Generative interfaces make that catalog a starting point. A music service can recommend a song or help make one for the occasion. A television interface can find a show or eventually let the viewer describe a scene, enter a familiar world, or alter what happens next.
Search becomes prompt. Recommendation becomes creation. The shelf starts to look like a conversation.
Measuring the moment
As those experiences change, so must the way we measure their value. We evaluate media through the economics of distribution: how many people watched, streamed, liked, or shared. Just-in-time media has a different job. If only four people hear a song, that is not a failure. It was made for, and partly by, those four people.
Reach will still matter, but it will tell us less about whether these experiences succeed. We will also need to ask whether someone got what they wanted from the moment and whether they return to make something for the next one. A song heard once can fulfill its purpose. A family making a new song every week can become a habit, even if no individual track attracts an audience.
This does not mean all entertainment will be just in time. We will still want talented people to tell great stories and write great songs. More creation will probably make trusted artists, recognizable worlds, and good curation more valuable.
But consumption platforms will need to support both behaviors: appreciating something finished and participating in what it can become. The winners will make just-in-time creation a natural part of consumption.
Tomorrow’s 1% rule may describe two different inequalities. Many more people will create something. A small percentage will still create what many other people choose to consume. What changes is how easily someone can move from consuming to creating, without seeking an audience.
The next media platform may not simply know what you want. It may invite you to create it.
And now, the world premiere of Bloomer:
